The average inflation here seems to be about 100-200 points, sometimes more sometimes less depending on how high up the ladder.  In addition to what's already been said, it may have something to do with the combination of the Glicko RD system and being able to hand-pick your online opponents, I'm not sure.  I think it's mostly understood among chess players that no rating system matches up with others, and not too many people are going around flaunting their chess.com rating as if it were a real qualification of chess skill.  Once the inflation has been adjusted a bit though, it will eliminate some confusion for sure.

I wish I knew what my exact otb rating was   It's been a while since I got any good estimate.

Have you ever played OTB Mirage?

Yes, a lot actually, but never in an official tournament.

I've sent you a game challenge.  Accept if you dare...!

Accepted, it doesn't matter to me if I lose, I'm not scared :)

Haha, you're CC rating is not like most CC ratings here... yours is a simul rating

Good answer.  I'm starting to wish again that there was a club I could go to where the average rating was 2200 and a good night would meaning winning or drawing one game (I don't live in New York heh).

Another thing to keep in mind (some vague pointers to this can be found in posts above) is that correspondence ratings can differ significantly from live play ratings. Jonathan Berry, for example, was an international grandmaster in the ICCF (long before competent chess computers were available), yet only reached FM in live play.

For blitz and bullet, I think that the ratings here are a bit deflated. For CC, a little bit inflated. Well, in both cases, I don't talk about cheaters (like the horde of 2500+ guys in CC). Even so, I found the ratings here pretty reliable: those who are better than me OTB have a higher rating, and the other way around. Sure, some small variations can occur (say from timeouts or the possibility to choose the opponents).

The problem you have is that you care about the rating as you consider it a measurement of your skill. You cannot accurately measure an online rating as the games are played under wildly different conditions. There is the family dad that squeezes in a move between fixing the toilet and bringing the kids to bed. There is the retiree that has all day to make a move. There is the basic member that is forced to move versus the paying member that has the advantage of being protected against timing out. There is the member that is on broadband who does not have to worry about internet access cost versus the dial-up member (I screwed up a few of my games while vacationing at my parents place as I felt forced to rush due to pay-by-the minute charges.

As others have said, the only accurate rating is OTB as the games are played under the (virtually) same conditions for everyone. Play is still useful. You have ample chances to improve your opening repertoire, hone your tactics and strategic thinking.

As for my inflated rating, at least I can say that my tactics trainer rating is in the same range as my current CC rating. So at least something is in sync. But until I play OTB again I really have no idea what my real rating is. However, I am sure it is a lot less than 2250.

Laudable goal, Erik. It is never simple though as there are some folks that prefer to timeout games rather than resigning them. In those cases, the winner gets shafted as he/she actually obtained a real win but would get denied the points due to the opponent timing out. I just came across those examples when I put together a list of individual team match performance for my group. Sigh ..

You mentioned elsewhere that your average thinking time is about 5 seconds per move. Orangehonda is right, you're playing a simul, not CC.

If you had an inclination to reduce your game load to 20, and spend 5 to 15 minutes per move, I'd be surprised if you did not get above 2600 on chess.com in very short time.

Good point about Glicko and hand-picking the opponents.

It should be fairly simple matter for chess.com staff to correlate the ratings reported in the user profiles and the chess.com ratings. Is this, where your estimate of 100-200 points comes from?

The only problem here is that a lot of people don't bother updating their ratings or even reporting them in the first place.

I mentioned the following data point in one of group forums, but perhaps it is worth sharing here too: a couple weeks ago, I looked at the rating differences of a number of players from the Bay Area where I am located -- mostly kids 7-14 years old and a couple adults, all below 2100 on chess.com, all below 1900 USCF. The average difference was a bit over 450 points. The differences appeared to diminish with the increasing USCF rating. Many players did not bother reporting their USCF ratings, and I got them only because I know who the players were.

However, this was a fairly small sample (about 20 players).

Maybe chess.com could poll the players to get a better handle on the correlation? Assuming there is nothing better to do, of course .

It's not clear on the survey if you're asking about cc or live.  Some people are answering according to live, come about cc, so when you ask the average amount of inflation or deflation you're getting two sets of answers under the same question.

Still, I like the idea.

Also think it might be helpful to have no established rating until 50 games are played, like USCF/FIDE does.  I noticed that there are people who abandon an account when their RD gets too low to move their rating much.

USCF provisional period is only 25 games (unless they've changed it)

I don't know all about FIDE established ratings, but I'm pretty sure it doesn't require 50.

*looks at my rating

I have done that a year ago with around 50 games, I went up to 2274. But there are too many cheaters out there, so 2350 would be a top for me now. GM Becerra is around 2600, so the gap between us should stay similar. That mean around 2300 for me.

A Chess.com rating, USCF rating and any other rating has the meaning you choose to give it.  What makes a USCF rating so special, I suppose, is that it's been around longer and it's harder to cheat when playing OTB.  So, if you are not cheating then the change in your rating on Chess.com is an accurate reflection of the change in your ability.  

The on-line changes in your rating may actually be a better reflection of your current strength because most play many more games on-line that OTB.  If you only play a few OTB tournaments a year, your USCF rating may take a very long time to catch up to the level of your current play particularly when you are using Tactics Trainer and watching the outstanding video lectures on Chess.com. Since becoming a paying member, my Live Chess rating had increased almost 300 points in well under a year.